Advanced visual and voice technology uses voice controlled AI solutions

    Renesas Electronics announced the joint development of a voice-controlled multi-mode AI solution that provides low-power contactless operations for image processing of the Internet of Things and edge systems based on visual AI, including self-checkout machines, security cameras and video conferencing systems. There are intelligent devices such as robot cleaning equipment.

    The new solution combines the company's RZ/V family of visual AI MPUs with the low-power multi-mode, multi-feature Syntiant NDP120 neural decision processor to provide advanced speech and image processing capabilities. The joint solution provides always-on functionality that quickly triggers voice activation from standby mode for target recognition, facial recognition and other vision-based tasks that are key features of security cameras and other systems. For example, user-defined voice prompts drive activation and system operations, while visual AI recognition tracks operator behavior and controls operations, or alerts when suspicious behavior is detected.

    The multimodal architecture makes it simpler to provide a touchless user experience for visual AI-based systems. Using a dedicated, energy-efficient speech recognition chip can reduce standby power consumption while speeding up system development, as it is feasible to develop software independent of visual AI capabilities.

    Hiroto Nitta, senior vice president and SoC business leader, said, "We expect the demand for multi-mode systems that use multiple input information streams, including images and voice, to increase as a way to improve ease of use and security." In Renesas Electronics' Internet of Things and Infrastructure business unit. "Through the partnership between Renesas Electronics, a leader in low-power imaging AI technology, and Syntiant, a leader in voice AI technology, we will accelerate the adoption of low-power, ultra-small intelligent voice AI technology in embedded systems and provide new integrated solutions for customers around the world."

    Kurt Busch, Syntiant's chief executive, said: "Voice-based user interfaces will make it possible for customers to deliver new user experiences, taking the next generation of innovation from concept to reality. "We have delivered over 15 million Deep learning NDP's worldwide, enabling always-on voice in a variety of consumer and industrial iot applications. Our partnership with Renesas delivers a powerful, low-power voice and imaging solution that is sure to accelerate our customer base in a variety of devices and use cases around the world."

    The new voice control multi-mode AI solution utilizes multiple compatible devices from Renesas's electronics portfolio, providing customers with an elevated prototyping platform to speed time-to-market and reduce risk. The new solution is part of the company's Winning Combinations, which offer a compelling combination of analog, power and embedded processing products to help customers speed their designs and get to market faster.